1. Field
The invention relates generally to the field of molecular diagnostic assays in microfluidic cartridges having on-board dehydrated reagents, where the reagents are stabilized by dehydration without lyophilization, and to methods of manufacture of microfluidic cartridges having same.
2. Description of the Related Art
Microfluidic cartridges are well suited for diagnostic molecular biology due to their convenience, operability by unskilled workers, and ready disposability. Each cartridge contains all reagents for a particular assay, so that only sample need be added. The cartridge is then simply inserted into an automated apparatus for performing the assay. However, since first conceived in the early 1990's, the road to development of these cartridges has taken two decades and obstacles to commercialization continue. Shelf-life of the cartridges is a particular concern, because most intended users lack access to frozen storage facilities.
While “freeze drying” has been successfully employed for extended storage of moisture-sensitive reagents at room temperature, the manufacturing of microfluidic cartridges does not readily permit use of freeze drying as part of the process. Additional assembly is required after the reagents are placed within wells or channels of the partially assembled device. Powders or loose spheres can become dislodged or fail to seat in the correct positions, and interfere with high speed assembly of cartridges formed from sheets or rolls. Dry reagents can also interfere with the use of adhesives. The final assembly of a microfluidic cartridge involves procedures for demasking, lamination, gluing and/or ultrasonic welding not readily compatible with freeze-drying technology on a commercial scale. Also, the moist or hot environments encountered in the assembly of microfluidic cartridges can inactivate freeze-dried reagents even before the product is fully assembled.
Alternatively, dehydration in glass form is known to preserve the function of enzymes or reagents during storage above freezing, but the art is highly unpredictable, and methods and compositions must be varied for each reagent studied—with no particular expectation of success.
Protein reagents that may be required for assembly of a self-contained microfluidic cartridge for a molecular biological assay include: TAQ polymerase, DNA polymerase, RNA polymerase, reverse transcriptase, RNAase H, proteinase K, immunoglobulin, luciferase, pyrophosphatase, chromopeptidase, lysozyme, and so forth. Other reagents likely to have sensitivity to light, moisture or heat include nucleotide triphosphates, deoxynucleotide triphosphates, primers and probes.
PCR forms the basis of a large family of molecular biological assays, and is currently the gold standard for many molecular diagnostics not available in a microfluidic cartridge format. Adapting PCR to a microfluidic format will involve developing methods for stable on-board storage of reagents including TAQ polymerase, reverse transcriptase, deoxyribonucleotide triphosphates, primers and probes. The DNA polymerases of thermophilic organisms used in PCR, generically referred to as “TAQ polymerases” by virtue of their first discovery in Thermus aquaticus, have proven particularly difficult to stabilize for room temperature storage. Nonetheless, the development of PCR products with a commercially useful shelf life at room temperature in a microfluidic device depends on a solution to this problem.
TAQ has 5′-3′ polymerase and exonuclease activity, but not the 3′-5′ proofreading capacity of other polymerases. The enzyme structure, however, is shared with other DNA polymerases and contains an opposable thumb-palm domain split by a deep cleft associated with DNA template binding. Features associated with thermostability of TAQ polymerases include increased ratios of Arg to Lys, Glu to Asp, Ala to Gly, Thr to Ser, and an absence of cysteine. Folding at elevated temperature is established and maintained by hydrophobic, hydrogen bonding, electrostatic and van der Waal's interactions.
Enzymes are complex folded nanomachines, having cooperative motions and flexibility related to their catalytic function and folding. Certain structural sub-domains are relatively fixed in structure and others are more fluid and dynamic. Ideally, the native state is preserved during storage by dehydration, but dehydration most commonly results in some level of destabilization of folding. Denaturation and loss of activity results from enzyme unfolding; changes in structure following dehydration (or freezing) can be so severe that refolding into an active “native state” form does not occur following rehydration.
The role of water in enzyme structure is firmly established. The degree of hydration of a protein may be expressed by “Dh”, where Dh≅0.4 (g H2O/g protein) indicates a full hydration shell or monolayer of water surrounding the protein. Intermediate levels of hydration are also known. At Dh≅0.15-0.2 water is sufficient only to associate with more polar and ionic surfaces and enzymatic activity is lost. Most lyophilization processes result in Dh≅0.02. In the absence of the dielectric shielding of water, electrostatic interactions can result in denaturation. Water dominates protein structure by continuously breaking and reforming hydrogen bonds in the hydration shell (leading to both hydrophobic and hydrophilic interactions), as well as by guiding secondary and tertiary structure such as α-helix and β-turn motifs through inter-peptide and side chain interactions. The liquid crystalline, hydrogen bonding capacity of water as a solvent lubricates or “plasticizes” the motions of structural domains of the enzyme.
Amorphous solids are preferred for “dry” storage of reagents because rehydration proceeds more rapidly than for the corresponding crystalline state. Ideally, the protein is stabilized in a solid, non-hygroscopic, glassy matrix, which undergoes a controlled devitrification when rehydrated with excess water. The preferred state has much in common with the glassy state formed by supercooling a liquid. Similarly, protein domains can be frozen in an amorphous “glassy” or gel-like state at or below a temperature Td (dynamical transition temperature), which is analogous to the Tg (glass transition temperature) for formation of a glassy state in small molecules and polymers. Below Td, protein unfolding is effectively inhibited. Similarly, dehydration to a critical level can be associated with inhibition of protein unfolding: at Dh<0.2 the hydration shell is patchy, and there are insufficient water molecules to execute the hydrogen bond rearrangements associated with unfolding of protein domains, even though the thermal energy available at room temperature is sufficient to denature the protein.
Of particular interest is the dehydration of proteins within glasses composed of lyoprotectants, molecules that protect the protein from denaturation during dry storage. Activity of lyoprotectants is perhaps best explained by a “water replacement model” in which the lyoprotectant is thought to interact directly with the protein through hydrogen and hydrophobic bonding, somehow offsetting the denaturing effect of removal of water. Glycerol, for example is thought to substitute for water in the protein's hydration shell and to effectively plasticize the dehydrated protein in a rehydratable form, albeit without the conformational instability of water.
Thus a common framework may be used to consider the amorphous glassy state formed by cooling a protein in an intimate mixture with a glass-forming molecule and the amorphous glassy state formed by dehydration of that mixture. The solid product in both cases is composed of protein conformers having varying degrees of native state which are “solvated” and molecularly dispersed in an amorphous glass such as a sugar. Protein and sugar mixtures for example have been found calorimetrically to have a bulk Tg intermediate between the Tg of the sugar and the Td of the protein in proportion to the composition of the mixture. Similarly, the Td of a protein may be modulated by intimate association of the protein with a suitable lyoprotectant, although the mechanism is not fully understood. Thus the conformation of the dewatered protein is believed to be somehow coupled to the molecular structure of the glass.
Candidate lyoprotectants include polyhydroxy compounds (PHCs) generally, particularly a variety of sugars (including monosaccharides, disaccharides, trisaccharides, and oligosaccharides), sugar alcohols, and a variety of polyhydroxy small molecules and polymers. Lactitol, mannitol, maltitol, xylitol, erythritol, myoinositol, threitol, sorbitol (glucitol), and glycerol are examples of sugar alcohols. Non-reducing sugars include sucrose, trehalose, sorbose, stachyose, gentianose, melezitose and raffinose. Derivatives of sugars that are lyoprotectants include methylglucoside and 2′-deoxyglucose, among others. Sugar acids include L-gluconate and metallic salts thereof. Less preferred for most applications include reducing sugars such as fructose, apiose, mannose, maltose, isomaltulose, lactose, lactulose, arabinose, xylose, lyxose, digitoxose, fucose, quercitol, allose, altrose, primeverose, ribose, rhamnose, galactose, glyceraldehyde, allose, apiose, altrose, tagatose, turanose, sophorose, maltotriose, manninotriose, rutinose, scillabiose, cellobiose, gentiobiose, and glucose. Also useful are polyvinylpyrrolidones, polyacrylamide, polyethylimine, pectin, cellulose, derivatized celluloses such as hydroxymethylcellulose, hydroxyethylcellulose, and hydroxypropylmethylcellulose, hydroxyethylstarch, soluble starches, dextrans, highly branched, high-mass, hydrophilic polysaccharides such as Ficoll®. Glass-forming albumins, gelatins and amino acids have also found use. By trial and error, useful mixtures of the above have also been discovered, typically differing for each target protein.
Success in formation of a glass is also known to be sensitive to rate of cooling, concentration, pressure and other process parameters such as the presence or absence of seed crystals. It must be recalled that a glass is a metastable state. The difficulties of these complex systems are illustrated by the following example, taken from WO 1996/033744, where it was reported that an amorphous solid freeze-dried composition of calcitonin 2% in lactose 95% with 2% residual water was raised above its Tg of 40° C., resulting in irruptive crystallization of the lactose and formation of a water of crystallization composed of 60% water and 40% protein, which was excluded from the crystalline phase. The glass temperature of the solution phase was below freezing and as a result the protein then very rapidly lost biological activity at room temperature. Similar inactivation of enzymes has been noted with crystallized sucrose (Schebor C et al, 2008, Glassy state and thermal inactivation of invertase and lactase in dried amorphous matrices. Biotech Progress 13:857-863).
Rosen, as disclosed in expired U.S. Pat. No. 4,891,319, discovered that trehalose, which has a higher Tg than lactose or sucrose, is lyoprotective when proteins are dried at room temperature, avoiding the rigorous conditions of freeze drying and spray drying, and reported that fluorescence markers may also be dehydrated in this way. Rosen suggested sugar:protein ratios of 1.4:1 to 10:1. Trehalose was proposed to act as a dry scaffold maintaining the structural integrity of the macromolecule when water was removed. These findings were further extending elsewhere (Colaco C et al, 1992, Extraordinary stability of enzymes dried in trehalose: simplified molecular biology, Bio/Technology 10:1007-11) and in U.S. Pat. No. 5,955,448 it was reported that various carbohydrates, including lactose or sucrose, may be employed as long as the formulation also includes an inhibitor of the Maillard browning reaction. Related observations have been reported by Franks (U.S. Pat. No. 5,098,893) and by Wettlaufer (U.S. Pat. No. 5,200,399), with comments on the importance of oxygen, light and chemical reactions in loss of activity of vitrified biological substances.
Sucrose, sorbitol, melezitose and raffinose have also been suggested as preferred lyoprotectants. However, to our knowledge, no success has been reported in stabilizing, without lyophilization, dry TAQ for extended storage stability periods with trehalose or any other sugar. To the contrary, in the declaration of A Madejón (FIG. 5—source: file wrapper of U.S. patent application Ser. No. 10/292,848), it is shown that trehalose is at most partially protective in dry reagent forms stored at 4° C. for 1 week, and that the standard PCR mixture without a lyoprotectant has more residual activity at 37° C. after one week than the dry reagent with the lyoprotectant. Madejón further shows that melezitose of itself is not protective at all. Referring to the gel, which is reproduced here as FIG. 5, lanes 1-9 (between the ladders) were run after reactant storage at 4° C.; lanes 10-18 after storage at 37° C. (“M”—melezitose, “L”—lysine, “G”—glycogen, “T”—trehalose, “S”—standard mix with no lyoprotectant).
Trehalose has been reported as unusual or even extraordinary in that addition of small amounts of water does not depress Tg, as in other sugars (Crowe J H et al, 1998, The role of vitrification in anhydrobiosis. Ann Rev Physiol 60:73-103). Instead, a dihydrate crystal of trehalose forms, thereby shielding the remaining glassy trehalose from effects of the added water. Franks, however, in U.S. Pat. No. 6,071,428 shows that this effect is not remarkable, and that raffinose pentahydrate is also useful in storing enzymes in a dry state. The crystalline pentahydrate is reported to coexist with a surrounding glassy state of anhydrous material. These hydrated saccharides are not generally associated with formation of waters of crystallization or irruptive crystallization which would favor denaturation.
Arieli, in WO 2007/137291, proposes stabilization of TAQ with stabilizing agents such as sucrose, trehalose, melezitose, sugar alcohols, and reducing sugars in combination with BSA by drying above freezing, typically by drying at 55° C. for 1-3 hrs. Qualitative data illustrated in the application demonstrate activity of TAQ after overnight or short term storage. However, no indication is given as to the degree of hydration (Dh) achieved in the drying process, and as is already known, TAQ retains full activity overnight in aqueous solution at room temperature (FIG. 6—source: Marenco A et al, 2004, Fluorescent-based genetic analysis for a microfluidics device, Defence R&D Canada Contract W7702-00-R849/001/EDM Final Report), and presumably for short term storage as well. Thus it is unclear whether the dehydration and glassy state achieved was sufficient for long term storage over months or years. Rosado (US Pat. Appl. 2003/0119042) has argued that TAQ is best stabilized in a fully hydrated “gelified” form, however the data disclosed suggests that only limited duration of stability was achieved, perhaps a few days or weeks.
Development of frozen commercial formulations of TAQ have been reported in U.S. Pat. No. 6,127,155, for example. However, frozen storage requires special equipment typically not available at point-of-care facilities where microfluidic cards find usage. Also of note, a number of investigators have reported success lyophilizing TAQ preparations. These include Walker (U.S. Pat. No. 5,565,318), Treml (U.S. Pat. No. 5,763,157), and Park et al. (U.S. Pat. Nos. 5,861,251 and 6,153,412). Park describes lyophilization of TAQ in the presence of glucose, sorbitol, sucrose or Ficoll®. Klatser P R et al describe a lyophilized PCR Mix using trehalose as cryoprotectant and Triton X-100. Klatser found TAQ activity of their lyophilized mixture when rehydrated at up to 1 year post preparation. Commercially available lyophilized beads containing TAQ with excipients are also available (Ready-to-Go PCR beads, Amersham Biosciences; Sprint™ Advantage®, Clontech, Mt View Calif.). Once lyophilized, the products are hygroscopic and sensitive to humidity and must be immediately sealed. The products apparently must also be held on ice during the rehydration process with ultra-pure water and subsequently prior to use, rendering their use in next-generation reagents-on-board microfluidic devices difficult if not impossible.
In contrast, next-generation microfluidic devices are configured so that use of ice or pure water during rehydration of reagents is not possible. The device reagents are typically rehydrated by sample or by an eluate prepared from the sample, for example by the method of Boom (U.S. Pat. No. 5,234,809). Thus, there is still a need in the art for a method of achieving ambient stabilization of DNA polymerase in the context of a PCR reagent mix that does not involve lyophilization and retains sufficient reliability over an extended storage time sufficient for practical commercialization of sensitive diagnostic assays in a microfluidic card.
Thus the disclosures to date do not apparently enable a formulation suitable for extended stable dry storage of TAQ polymerase without lyophilization or freezing. As commercialization of microfluidic devices for diagnostic applications moves closer to fruition, a workable solution to this problem is more urgently needed.